This proposal describes a research plan and training program to develop the principal investigator, Dr. Laura Wood, from junior faculty into an independent physician-scientist. Dr. Wood has performed seminal work in cancer genomics and has completed clinical training in Gastrointestinal, Liver, and Pancreatic Pathology. She was appointed Assistant Professor in July 2013, a tenure-track position with 75% protected time for the proposed studies. Her long-term goal is to develop an independent research program to characterize the molecular events that underlie multistep tumorigenesis in the pancreas. This program will leverage her existing expertise in genomics and pancreatic pathology; however, it will also require additional training in bioinformatics and cell biology in order to independently identify and functionally validate molecular events in pancreatic neoplasms. In order to realize these goals, Dr. Wood has assembled a diverse mentorship team with expertise in cancer genomics, pancreatic pathology, cell biology, and bioinformatics. Her career development plan includes formal coursework as well as seminars, lab meetings, and hands-on training. The proposed research and career development will guide Dr. Wood to a position leading a productive independent research group. Invasive pancreatic cancer arises from non-invasive precancerous lesions that are curable if detected early enough. These precursors include low-grade lesions (which have a low risk of malignant transformation) as well as high-grade lesions (which are equivalent to carcinoma in situ). Early detection of high-grade precursors before their transformation to carcinoma is the key to improving survival of patients with pancreatic neoplasms. In spite of the clinical importance, little is known about the alterations that distinguish low-grade and high-grade precursors in the pancreas. The overall goal of these studies is to determine the somatic mutations that distinguish low-grade from high-grade precursor lesions in the pancreas. After careful pathological review of human tissue samples, we will perform next generation sequencing on matched low-grade and high-grade precursors from patients without associated invasive carcinoma. These studies will allow us to compare the genetic alterations in both types of precursors and identify somatic mutations that are specific to high-grade lesions. In order to distinguish drivers of the progression to high-risk lesions, we will validate the functional effects of these mutations via genetic manipulation in an innovative three-dimensional organoid culture model of human pancreatic neoplasia. Taken together, the proposed studies will help to identify the mutations that causally contribute to the formation of high-risk precursor lesions in the pancreas ? these changes represent the best targets for early detection, treatment, and prevention of pancreatic cancer.